Multiscale molecular modeling and simulation of carbon nanotubes

Abstract

Due to their excellent mechanical properties, carbon nanotubes (CNTs) have undergone intense scrutiny by the research community at large since their discovery in 1991. In this thesis, we propose and implement an overlapping scheme to concurrently couple empirical molecular dynamics (MD) with both semi-empirical tight-binding (TB) and ab initio density-functional theory (DFT). The multiscaling of these three approaches is then used to numerically study the mechanical properties and dynamic characteristics of CNTs. Since MD is based on empirical potential functions, it is relatively "cheap" in terms of computational intensity, and is therefore the technique of choice throughout most of the domains in the CNT during simulation. However, when there is large deformation and/or bond breaking, more robust and refined techniques are required, namely TB or DFT. Basically, the TB technique is a simplification of the ab initio method, where a minimal basis set is used in the Hamiltonian system.